Abstract: ‍ ‍The radar detection and recognition of low-slow and small targets， such as unmanned aerial vehicle（UAV） in complex environments， encounter significant challenges， while related research has become an important and difficult issue in the field of radar detection. The dataset is the basis of radar target detection and recognition research， and data quality and diversity play an important role in the performance validation of data processing algorithms. In the current publicly released dataset， radars are often deployed on the ground， with a relatively clear background for aerial target detection. At present， there are fewer publicly released radar UAV target-detection datasets under clutter background， while the detection scenarios， observation viewpoints， target flight heights， and signal bandwidths are relatively single. Therefore， data diversity needs to be improved. This study constructs a set of radar low-slow-small UAV detection datasets （LSS-Ku-1.0） under a ground-clutter background. Using a Ku-band phased-array radar placed on a high tower， the radar echo data of UAV targets with strong clutter background were recorded in the complex and wild environment of a jungle and grassland. The dataset covers target echoes with different signal waveforms， bandwidths， grazing angle variations， and three different flight altitudes. Based on the dataset， the statistical distribution and time-dependent characteristics of the clutter were analyzed. Five statistical models were used to fit the statistical distribution of clutter， and the results of goodness-of-fit tests were given. The micro-Doppler characteristics of the rotor blades of the UAV were analyzed. The high-resolution range profile， spectrogram， time-frequency map， and range-Doppler spectrum of typical data were also investigated. These provided data support for the analysis of radar characteristics of low-slow-small targets， as well as research on detection， tracking， and recognition.